Emotion-Synced LED Costume Display Project

ISEF Category: Technology Enhances the Arts

Subcategory: Display Technology  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A costume can feel alive before the actor says a word. Light, motion, and sound can push an audience’s emotions fast. Your project tests whether a breathing LED pattern actually changes how people feel, or if random flicker works just as well.

What Is It?

This project combines wearable lighting, speech analysis, and biosignals. The costume uses WS2812B LEDs, which are addressable lights you can control one by one, behind translucent fabric. The result can look like the costume is breathing, pulsing, or reacting to the performer.

The core idea is simple. You collect two signals from the actor, heart rate and voice features, then feed them into a model that predicts a lighting pattern. Heart rate is a body signal that often changes with stress or excitement. Vocal prosody means the rise and fall of speech, like pitch, loudness, and rhythm. Your question is whether people in the audience feel more emotional arousal when the costume follows those signals instead of flashing randomly.

Think of it like visual music. A plain light pattern is like a metronome. A signal-driven costume is like an instrument that listens back. Your job is to test whether that extra connection changes the viewer’s response.

Why This Is a Good Topic

This is a strong science fair topic because you can test a real design question with real human response data. You are not just making a cool costume. You are measuring whether responsive lighting changes audience emotion compared with a control pattern. That gives you a clear independent variable, a clear outcome, and room for analysis. You can also learn signal processing, machine learning, user studies, and basic experimental design.

Research Questions

• How does a heart rate synced lighting pattern change audience emotional-arousal ratings compared with random flicker?
• What is the effect of matching LED pulse timing to vocal prosody on perceived intensity of a performance?
• Does the addition of both heart rate and voice features improve audience ratings more than either signal alone?
• To what extent does translucent fabric opacity change how viewers interpret the same LED pattern?
• Which lighting feature, pulse speed, brightness change, or color shift, best predicts audience arousal ratings?
• How does a trained transformer model compare with a simple rule-based controller for creating emotionally matched light patterns?

Basic Materials

• WS2812B LED strip or panel with controller board.
• Translucent fabric samples with different weave or opacity.
• Microcontroller such as Arduino, ESP32, or similar.
• MAX30102 heart rate sensor module.
• Microphone for recording voice samples.
• Laptop for data collection and model testing.
• Smartphone or tablet for audience rating surveys.
• Tripod or stand for keeping costume and camera position stable.
• Power supply sized for the LED strip.
• Consent forms and survey sheets for human participant testing.

Advanced Materials

• WS2812B LED strips with diffuser materials and spare segments.
• MAX30102 sensor module with validated pulse logging setup.
• High-quality microphone and audio interface for voice capture.
• Computer with Python environment and GPU access if available.
• Motion capture or camera system for syncing performance timing.
• Light meter or colorimeter for calibration checks.
• Multichannel data logger for synchronized physiological and audio streams.
• Prototype wearable frame or garment base for repeatable mounting.
• Access to human subjects review guidance from a school or university.
• Stat analysis software for mixed-effects or repeated-measures testing.

Software & Tools

• Python: Processes sensor data, trains models, and compares lighting conditions.
• Whisper: Transcribes speech and helps extract vocal timing features.
• ImageJ: Measures fabric brightness and light spread from photos or video frames.
• Google Forms: Collects audience emotional-arousal ratings quickly and consistently.
• GeoGebra: Helps inspect trends, fit curves, and compare simple models.

Experiment Steps

1. Define the single lighting behavior you want to test, such as breathing, pulsing, or flicker, and decide what counts as a matched pattern.
2. Choose the audience response metric first, then build your survey so you measure arousal, not just general liking.
3. Plan a control condition that keeps the same costume, performer, and setting while changing only the light logic.
4. Decide how you will sync the heart rate, voice features, and LED timing so the system responds in a repeatable way.
5. Build a small pilot dataset to check whether the model separates useful signal from noise before you test viewers.
6. Set up a data analysis plan that compares conditions across viewers and checks whether differences are bigger than random variation.

Common Pitfalls

• Using inconsistent room lighting, which changes how bright the costume looks from one trial to the next.
• Training on too little voice and heart rate data, which makes the model react to noise instead of performance features.
• Picking a control condition that looks weaker by design, which makes the comparison unfair.
• Letting the performer change posture or movement between trials, which confounds the effect of the light pattern.
• Asking vague audience questions, which produces ratings that do not map cleanly onto emotional arousal.

What Makes This Competitive

A stronger project goes beyond making the costume work. You compare multiple controllers, not just one, and you measure whether viewers can tell the difference. You also need careful synchronization, clean controls, and a real statistical plan for repeated viewers or repeated performances. If you add a novel feature set, such as voice rhythm, pulse variability, and fabric diffusion together, your project becomes much more than a demo.

Project Variations

• Test the same emotion-synced lighting idea on a dance costume instead of a spoken-performance costume.
• Compare translucent fabrics with different opacities to see how material choice changes emotional response.
• Replace the transformer model with a simple rule-based timing system and compare viewer ratings.

Learn More

• PubMed: Search for review articles on physiological synchrony, human emotion, and audience response to performance cues.
• NIH RePORTER: Look up funded projects on wearable sensing, biosignals, and human-computer interaction.
• NOAA Education: Find clear explainers on signal processing and data interpretation methods that translate well to student projects.
• NASA Open Science: Search for open resources on time-series analysis and pattern recognition used in sensor data work.
• MIT OpenCourseWare: Browse free materials on machine learning, signal processing, and probability for the technical background.
• IEEE Xplore: Search for peer-reviewed papers on wearable displays, emotion-aware systems, and interactive textiles through school access or free abstracts.